three genes of a motility operon and their role in ... · a motility (mot) operon containing three...

JOURNAL OF BACTERIOLOGY,0021-9193/97/$04.0010

Oct. 1997, p. 6391–6399 Vol. 179, No. 20

Copyright © 1997, American Society for Microbiology

Three Genes of a Motility Operon and Their Role in FlagellarRotary Speed Variation in Rhizobium meliloti

JOSEF PLATZER, WERNER STERR, MARTIN HAUSMANN, AND RUDIGER SCHMITT*

Lehrstuhl fur Genetik, Universitat Regensburg, D-93040 Regensburg, Germany

Received 2 June 1997/Accepted 14 August 1997

The peritrichous flagella of Rhizobium meliloti rotate only clockwise and control directional changes ofswimming cells by modulating flagellar rotary speed. Using Tn5 insertions, we have identified and sequenceda motility (mot) operon containing three genes, motB, motC, and motD, that are translationally coupled. ThemotB gene (and an unlinked motA) has been assigned by similarity to the Escherichia coli and Bacillus subtilishomologs, whereas motC and motD are new and without known precedents in other bacteria. In-frame deletionsintroduced in motB, motC, or motD each result in paralysis. MotD function was fully restored by complemen-tation with the wild-type motD gene. By contrast, deletions in motB or motC required the native combinationof motB and motC in trans for restoring normal flagellar rotation, whereas complementation with motB or motCalone led to uncoordinated (jiggly) swimming. Similarly, a motB-motC gene fusion and a Tn5 insertionintervening between motB and motC resulted in jiggly swimming as a consequence of large fluctuations inflagellar rotary speed. We conclude that MotC biosynthesis requires coordinate expression of motB and motCand balanced amounts of the two gene products. The MotC polypeptide contains an N-terminal signal sequencefor export, and Western blots have confirmed its location in the periplasm of the R. meliloti cell. A workingmodel suggests that interactions between MotB and MotC at the periplasmic surface of the motor control theenergy flux or the energy coupling that drives flagellar rotation.

Motile bacteria swim by rotating helical flagellar filamentsdriven by a rotary motor in the cell membrane that is energizedby the transmembrane gradient of protons (37, 41) or by so-dium ions (20). The structure of bacterial flagella, their genet-ics, and the physiology of flagellar rotation have been exten-sively studied in members of the g subgroup of proteobacteria,such as Escherichia coli and Salmonella typhimurium (for re-views, see references 22, 35, and 54).

In E. coli, the assembly, structure, and function of flagellarequire the products of some 50 genes. Among these, themotility proteins, MotA and MotB, and the switch proteins,FliM, FliN, and FliG, are necessary for flagellar rotation andfor alternating between clockwise and counterclockwise rota-tion (30, 35). The membrane-bound proteins, MotA andMotB, are essential for torque generation (2–4, 63, 65). Theseproteins are known to function in transmembrane proton con-duction, and MotB may also serve as an anchor that attachesthe stator (the nonrotating portion of the motor) to the pep-tidoglycan layer (7, 9, 26, 61, 67). MotA-MotB complexes prob-ably correspond to the 11 or 12 “studs” seen in electron mi-crographs as rings of intramembrane particles encircling theflagellar rod (24). Furthermore, averaged electron cryomicro-graphic images of the basal body have revealed a bell-shapedstructure, named the C ring, that is attached to the cytoplasmicface of the basal body (13) and contains the switch proteins,FliM, FliN, and FliG, essential for flagellar rotation and direc-tion control (21, 25, 36, 59).

Rhizobium meliloti, a member of the a subgroup of pro-teobacteria (45), exhibits distinctive differences from the en-terobacterial paradigm (15, 16). An R. meliloti cell has typically5 to 10 peritrichously inserted complex flagella, their filamentsconsisting of four related flagellin subunits (47, 56). The com-

plex flagellar filaments form right-handed helices capable oflimited polymorphic transitions but, as a result of helical ridgeson their cylindrical surface, are unable to switch handedness(66). Accordingly, flagellar rotation in R. meliloti is unidirec-tionally clockwise, and swimming cells respond to tactic stimuliby modulating the flagellar rotary speed (60). Rotary speed iscontrolled by two response regulators, CheY1 and CheY2, andit has been postulated that their phosphorylation by CheAautokinase acts to slow the flagellar motor (60).

Both unidirectional rotation and speed variation are newfeatures that prompted this study of the genetic basis of mo-tility in R. meliloti. We report here the analysis of three closelylinked motility genes, motB, motC, and motD, forming the motoperon. Whereas the unlinked motA gene (62) and motB areanalogs of two familiar enterobacterial mot genes, motC andmotD encode additional functions related to the new mode offlagellar rotation.

MATERIALS AND METHODS

Bacterial strains and plasmids. Derivatives of E. coli K-12 and R. melilotiMVII-1 (23) and plasmids used in this study are listed in Table 1.

Media and growth conditions. E. coli strains were grown in LB medium (34)at 37°C. R. meliloti strains were grown in TYC (0.5% tryptone, 0.3% yeastextract, 0.13% CaCl2 z 6H2O [pH 7.0]) or in RB minimal medium supplementedwith vitamins and 0.2% mannitol as the sole carbon source (14) at 30°C. Agarplates were solidified with 1.5% Bacto Agar (Difco). Single colonies were trans-ferred by toothpick onto swarm plates containing RB salts, 1024 M mannitol, and0.3% agar and incubated at 30°C for 2 days. Tn5 mutagenesis was conducted aspreviously described (16). For optimum numbers of swimming cells, R. meliloti(108 cells/ml) was starved overnight in RB medium without mannitol, sedimentedby centrifugation at room temperature, resuspended in an equal volume of RBmedium with 0.2% mannitol, and grown in a rotary shaker (New BrunswickScientific Co., New Brunswick, N.J.) at 30°C to 3 3 108 cells per ml (about 4 h).

Cloning of motility genes. Tn5-tagged chromosomal EcoRI fragments of mu-tant strains RU11/112 and RU11/113 were cloned into pIC20R and used totransform E. coli JM109 as described by Sambrook et al. (51). Transformantscontaining a Tn5 insertion were selected on LB agar plates supplemented withampicillin (100 mg per ml) and kanamycin (40 mg per ml). Adjacent fragmentswere obtained by partial EcoRI digestion of RU11/113 DNA. The resultingplasmids, pRU1213, pRU1212, pRU2064, and pRU2005 (Table 1), were used forsequence analysis.

* Corresponding author. Mailing address: Lehrstuhl fur Genetik,Universitat Regensburg, D-93040 Regensburg, Germany. Phone: 49(941) 9433162. Fax: 49 (941) 9433163. E-mail: [email protected].

6391

on January 25, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

DNA methods. Chromosomal DNA from R. meliloti was extracted from ly-sozyme-treated cells and purified by dye-buoyant density centrifugation as pre-viously described (48). For PCR, chromosomal DNA was extracted from singlecolonies of R. meliloti taken from an agar plate. Colonies were suspended in 100ml of distilled water mixed with glass beads and one aliquot of phenol, agitatedon a Vortex mixer for 1 min, incubated at 65°C for 5 min, and mixed, and debriswas sedimented at 14,000 rpm for 5 min. The aqueous phase was extracted twicewith one aliquot of chloroform; the DNA was ethanol precipitated and used forPCR. Plasmid DNA for sequence analysis was purified by the method of Li andSchweizer (32) optimized for GC-rich DNA templates, and PCR-amplified DNAwas cloned by DISEC-TRISEC techniques (11). DNA was sequenced as de-scribed by Sanger et al. (52), using a Quick Denature Sequenase kit fromAmersham Buchler (Braunschweig, Germany) and the supplier’s protocol.

Expression of reMotC and antibody preparation. Recombinant MotC (re-MotC) proteins, either complete (starting at Met1) or truncated (starting atMet36 [see Fig. 3]) were obtained by ligation to a 59 leader sequence encodinga His tag positioned under the isopropylthiogalactopyranoside (IPTG)-inducibleT7 promoter of pET19B and overexpressed in E. coli BL21(DE3) (50, 64). Lowyields of complete reMotC and high yields (approximately 15 mg from ca. 1011

E. coli cells) of truncated reMotC (lacking the hydrophobic signal sequence)were obtained and purified by Ni-nitrilotriacetic acid affinity chromatography(Qiagen) to near homogeneity as monitored by sodium dodecyl sulfate-poly-acrylamide gel electrophoresis. Polyclonal antibody against purified truncatedreMotC was prepared from rabbits as described previously (40) or producedcommercially (Eurogentec, Seraing, Belgium). The crude antiserum (0.5 ml) waspurified by sequential depletion against 1-ml extracts of E. coli BL21(DE3) andthe MotC-deficient R. meliloti RU11/513 (Table 1), each at 4°C for 60 min.Precipitating material was sedimented (16,000 3 g, 30 min, 4°C), and the super-natant was run through a Sepharose-protein A column (Pharmacia, Freiburg,Germany). Antibody contained in the eluate was further purified by chromatog-

raphy on reMotC bound to CNBr-activated Sepharose (Pharmacia). Aliquots ofthe purified antiserum were stored at 220°C and used at 1:200 dilution.

Cell fractionation. For membrane preparations, approximately 5 3 1010 mo-tile R. meliloti cells were harvested by centrifugation (5,000 3 g, 5 min), washedtwice in 1 volume of 100 mM Tris-HCl (pH 8.0) at 4°C, and resuspended in 80ml of 30 mM Tris-HCl (pH 8.0) containing 20% sucrose. After addition of 10mM K2-EDTA (pH 7.0) and 0.5 mg of lysozyme (final concentrations), proto-plasts formed during a 30-min period of gentle stirring at room temperature.These were sedimented by centrifugation (16,000 3 g, 10 min, 4°C) and resus-pended in 0.5 ml of lysis buffer (0.1 M KPi [pH 6.6], 20% sucrose, 20 mMMgSO4), and 5 mg each of DNase and RNase were added. Protoplasts were lysedby dilution in 80 ml of KPi (pH 6.6) and incubation at 37°C for 15 min undervigorous agitation; the mixture was adjusted to 10 mM K2-EDTA (pH 7.0) and15 mM MgSO4 and incubated again for 15 min. Cell debris was collected bycentrifugation (16,000 3 g, 30 min), and the supernatant containing membraneswas sedimented (45,000 3 g, 12°C, 30 min). The pellet was resuspended in buffer(100 mM KPi [pH 6.6], 10 mM EDTA), and the procedure was repeated four tosix times to yield pure membranes. The periplasmic fraction was prepared by thepolymyxin method (31, 39). About 5 3 1010 motile cells were harvested (4,000 3g, 10 min, 4°C), washed with 1 ml of buffer A (50 mM NaCl, 2 mM morpholine-propanesulfonic acid [MOPS] [pH 7.5]), sedimented again (5,000 3 g, 10 min,4°C), and resuspended in 600 ml of buffer B (10 mM MOPS, 5 mM EDTA, 5 mMPefabloc [Merck, Darmstadt, Germany]) containing 2 mg of polymyxin B (Sigma,Munich, Germany) per ml. Polymyxin B permeabilizes outer membranes ofgram-negative bacteria without affecting the cytoplasmic contents, thereby re-leasing the periplasm into the medium (31). After 60 min of incubation at 4°C,cells were sedimented (16,000 3 g, 30 min, 4°C). Centrifugation of the superna-tant (400,000 3 g, 120 min, 4°C) yielded a small pellet that was carefully resus-pended in 20 ml of H2O and used for immunoblot analysis. The cytoplasmicfraction was prepared from polymyxin B-treated cells after harvesting (5,000 3

TABLE 1. Bacterial strains and plasmids used

Strain or plasmid Markersa or derivation Reference or source

E. coliJM109 recA1 endA1 gyrA96 thi hsdR17 supE44 relA1 D(lac proAB)/F9 traD36 proAB lacIqZDM15 68S17-1 recA endA thi hsdR RP4-2-Tc::Mu::Tn7 Tpr Smr 58BL21(DE3) F2 ompT lon (rB

2 mB2) 64

R. melilotiRU11/001 Smr; spontaneous streptomycin-resistant derivative of RU10/406 (wild type) 27RU11/112 Smr Neor; Tn5 insertion derivative of RU11/001, uncoordinated (jiggly) swimming This study (Fig. 2)RU11/113 Smr Neor; Tn5 insertion derivative of RU11/001, paralyzed (Mot2) This study (Fig. 2)RU11/211 Smr; deletion in motC, paralyzed (Mot2) This study (Fig. 2)RU11/212 Smr; deletion in motD, paralyzed (Mot2) This study (Fig. 2)RU11/213 Smr; fusion of motB and motC (jiggly) This study (Fig. 3)RU11/218 Smr, deletion in motB, paralyzed (Mot2) This study (Fig. 2)RU11/215 Smr Gmr Neor; RU11/211 (pRU2014) This studyRU11/216 Smr Gmr Neor; RU11/212 (pRU2015) This studyRU11/219 Smr Gmr Neor; RU11/218 (pRU2016) This studyRU11/513 Smr; motB and motC, in-frame deletion, paralyzed (Mot2) This study (Fig. 2)RU11/514 Smr Gmr Kmr; RU11/513 (pRU1902) This study

PlasmidspIC20H/R Apr 38pUCBM20 Apr Boehringer MannheimpML122 Gmr 29pET19B Apr 64pK18mobsacB Kmr 53pRU1212 Apr Kmr; in vitro recombinant of pIC20R and 3.8-kbp EcoRI fragment of R. meliloti

RU11/112 with Tn5 inserted between motB and motCThis study

pRU1213 Apr Kmr; in vitro recombinant of pIC20R and 3.8-kbp EcoRI fragment of R. melilotiRU11/113 with Tn5 inserted in motC

This study

pRU1902 Gmr Kmr; in vitro recombinant of pML122 and 3.0-kbp fragment containing motB andmotC

This study

pRU2005 Apr Kmr; in vitro recombinant of pIC20R and 7.7-kbp EcoRI fragment of R. melilotiRU11/113

This study

pRU2014 Gmr Kmr; in vitro recombinant of pML122 and 1.3-kbp fragment containing motC This studypRU2015 Gmr Kmr; in vitro recombinant of pML122 and 1.6-kbp SfuI-BamHI fragment containing

motDThis study

pRU2016 Gmr Kmr; in vitro recombinant of pML122 and 1.2-kbp PCR-generated fragmentcontaining motB

This study

a Nomenclature according to Bachmann (1) and Novick et al. (44).

6392 PLATZER ET AL. J. BACTERIOL.


.org/D

ownloaded from

http://jb.asm.org/

g, 5 min) and resuspension in 1 volume of 50 mM Tris (pH 6.6)–10 mM EDTAby three passages through a French press at 10,000 lb/in2. The resulting extractwas freed of cell debris by centrifugation (45,000 3 g, 30 min). Total cell extractswere obtained by the same procedure without polymyxin B treatment, startingfrom 5 3 1010 intact motile cells.

Alkaline phosphatase and b-galactosidase served as marker enzymes for theperiplasmic and cytoplasmic fractions, respectively (33). Alkaline phosphataseactivity was determined by the hydrolysis of p-nitrophenyl phosphate (Sigma) asdescribed by Brickman and Beckwith (6), and b-galactosidase was assayed byhydrolysis of o-nitrophenyl-b-D-galactopyranoside (Sigma) as described by Miller(42).

Western blotting. Samples containing 30 to 50 mg of protein were separated ona 0.1% sodium dodecyl sulfate–12% polyacrylamide gel at 30 mA for 90 min,transferred to nitrocellulose, treated with 1% blocking reagent (BoehringerMannheim) for 60 min, washed once with 0.5% blocking reagent, and thenincubated overnight at room temperature with anti-reMotC antiserum (1:200dilution). The nitrocellulose membrane was washed twice with TBS-T (20 mMTris-HCl, 138 mM NaCl, 0.2% Tween 20) and 0.5% blocking reagent for 15 mineach time. Bands were visualized by chemiluminescence using goat anti-rabbitantibody (1:2,000) complexed with peroxidase and an enhanced chemilumines-cence detection kit (Boehringer Mannheim).

Deletion by allele substitution. Deletions were generated in vitro by using twoPCR steps as described by Higuchi (19). PCR products containing the desireddeletion were cloned into the mobilizable suicide vector pK18mobsacB (53)and used to transform E. coli S17-1. Filter crosses of E. coli and R. melilotiwere performed as described by Simon et al. (58). Exconjugant R. meliloticells were selected on TYC agar supplemented with neomycin (120 mg perml) and streptomycin (600 mg per ml) for counterselection of E. coli at 30°C.Single colonies of R. meliloti (containing cointegrated plasmid DNA) weregrown overnight in TYC to allow resolution and homologous recombinationto occur. Selection on RB agar supplemented with 20% sucrose (55) at 30°Cyielded colonies that had lost the sacB-containing vector. These werescreened for neomycin sensitivity and for swarming proficiency. All coloniesexhibiting unusual swarming behavior (approximately 1 in 200) were moni-tored by PCR and Southern hybridization to determine whether the mutantphenotype resulted from a directed deletion in the target gene. Complemen-tation with pML122-based recombinant genes (Table 1) was performed asdescribed by Labes et al. (29).

Motion analysis. Motility and swimming patterns of R. meliloti were observedby phase-contrast microscopy (Zeiss Standard 14 microscope; Zeiss, Oberko-chen, Germany). Free-swimming tracks were analyzed by computerized motionanalysis using a Hobson BacTracker (Hobson Tracking Systems, Ltd., Sheffield,England) as previously described (60).

Nucleotide sequence accession number. The sequence shown in Fig. 3 hasbeen deposited in GenBank under accession no. L49337.

RESULTS

Transposon mutagenesis. Insertion mutagenesis and taggingwith transposon Tn5 were used to identify genes encodingcomponents of the R. meliloti flagellar motor. Of 8,000 coloniesscreened on swarm plates, two motility mutants, RU11/112 andRU11/113, were isolated. These were examined electron mi-croscopically and were found to possess intact flagella. StrainRU11/113 was nonmotile on swarm plates (Fig. 1A) and oninspection by phase-contrast microscopy. By contrast, strainRU11/112 formed small swarm rings (Fig. 1A), and motionanalysis revealed cells that rotated (jiggled) around their axisor swam in tiny spirals (Fig. 1C). If tethered by a single flagel-lum, single unidirectionally rotating cells exhibited extremefluctuations of rotary speed (data not shown).

Physical map. Southern hybridization of genomic DNAfrom mutants RU11/112 and RU11/113 detected single trans-poson insertions. Chromosomal fragments obtained afterEcoRI digestion of RU11/112 and RU11/113 DNA werecloned into plasmid pIC20R and then used to transform E. coliJM109. Tn5-containing fragments identifiable by their resis-tance to kanamycin were physically mapped. The paralyzedmutant RU11/113 and the jiggly mutant RU11/112 containedTn5 inserted into identical 3.8-kbp EcoRI fragments. Sequenceanalysis (below) confirmed that the two Mot phenotypes rep-resent two alleles of a new motility gene, named motC. Achromosomal walk from the transposition sites led to the iden-tification of a 7.7-kbp genomic region including three closelylinked open reading frames (Fig. 2). These were assigned asmot on the basis of in-frame deletions introduced in motB(strain RU11/218), motC (strain RU11/211), and motD (strainRU11/212) each resulting in paralyzed cells with intact flagella.Translational coupling of the three genes (46), the interdepen-dence of the encoded functions (shown below), and commonupstream promoter and downstream terminator sequences(Fig. 3) all suggest that motB-motC-motD form an operon. Thephysical map (Fig. 2) depicts gene sizes and their polarity, Tn5insertions resulting in two different MotC phenotypes, and the

FIG. 1. (A) Swarm test of wild-type R. meliloti RU11/001 (swarm 1), the Tn5-induced jiggly mutant RU11/112 (swarm 2), and the Tn5-induced paralyzed mutantRU11/113 (swarm 3). Strains to be tested were transferred by toothpick onto swarm plates containing 10 mM mannitol and incubated at 30°C for 2 days. The diameterof a swarm ring reflects the motile proficiency of the respective strain. (B and C) Swimming paths of wild-type R. meliloti RU11/001 (B) and jiggly mutant RU11/112(C). Swimming cells were monitored by computerized motion analysis.

VOL. 179, 1997 MOTILITY GENES IN R. MELILOTI 6393


.org/D

ownloaded from

http://jb.asm.org/

extent of in-frame deletions introduced to define gene func-tions.

Nucleotide sequence and gene assignments. The 4,196-bpnucleotide sequence that contains the mot operon and flankingDNA is presented together with deduced polypeptide se-quences in Fig. 3. Gene assignments were made by using thesimilarity to the E. coli homolog (motB) and by phenotypes ofdeletion mutants delineated in Fig. 2. The sequence of a motAhomolog encoding a protein of 63% similarity to the E. coliMotA and mapping 21 kbp upstream on the R. meliloti chro-mosome has been deposited in a data library (GenBank acces-sion no. U87913) and will be described elsewhere. Closelinkage of the other three mot genes—motB extends fromnucleotides 202 to 1387 (394 codons), motC extends from nu-cleotides 1390 to 2694 (434 codons), and motD extends fromnucleotides 2691 to 4118 (475 codons)—indicates operonstructure, and the juxtaposition of genes with a putative ribo-some-binding site overlapping the upstream open readingframe (Fig. 3) suggests translational coupling (46). A s28-likepromoter motif (18) upstream of motB and a short terminatorpalindrome followed by thymidine residues downstream ofmotD are marked as candidate transcription signals for a poly-cistronic mRNA (Fig. 3). Yet, Northern analysis and the un-ambiguous assignment of transcription initiation by primerextension were obviated by low-level mot mRNA.

Functional assignments. (i) MotB. The derived 394-residueMotB polypeptide of R. meliloti exhibits 34% identity and 55%similarity to the E. coli MotB (61) but is distinctly longer thanthose of E. coli (308 amino acid residues [61]) and Bacillussubtilis (261 residues [43]), due to a large insertion betweenresidues Gln71 and Ser204 (Fig. 4). On the other hand, regionsessential for function are well conserved: the hydrophobic re-gion between residues Lys33 and Ala54 closely resembles theproposed membrane-spanning a helix of E. coli MotB (7, 61),with a conserved Asp37 thought to participate in transmem-brane proton conduction together with charged residues offour a-helical domains of MotA (4, 8). Furthermore, a pepti-doglycan-binding motif, identified in E. coli and B. subtilis

MotB proteins and in OmpA-related outer membrane proteins(9, 26), is conserved in the R. meliloti MotB protein (Asn335 toLeu350 and aligned consensus motif in Fig. 4). It has beenproposed that the association of this domain with the pepti-doglycan layer serves to anchor the MotA-MotB complex tothe cell wall (7). Moreover, residues that affect the function ofMotB in E. coli (5) (highlighted in Fig. 4) are highly conserved.

Complementation of the paralyzed motB deletant RU11/218with an intact motB overexpressed under neomycin promotercontrol (strain RU11/219) restored less than 30% of swarmingproficiency (Fig. 5B). Motion analysis of free-swimming RU11/219 cells revealed considerable reductions in absolute (223%)and relative (240%) speed and hence a low smooth-swimmingrate (Table 2). This was similarly observed when the paralyzedmotC deletant RU11/211 was complemented with an intactmotC gene (RU11/215): swarming (Fig. 5C) as well as absolute(218%) and relative (252%) free-swimming speed was greatlyreduced, and so was the smooth-swimming rate (Table 2).Likewise, overexpression of a plasmid-borne motB or motCgene in wild-type R. meliloti RU11/001 had an adverse, semi-dominant effect on motility, very similar to that seen in theDmotB mutant complemented with an intact motB gene(DmotB/motB) and DmotC/motC complementation experi-ments (Fig. 5 and Table 2). The values range in between thoseof wild-type RU11/001 and the jiggly mutant RU11/112, be-cause the merodiploids always formed mixed populations ofnormal and jiggly swimmers.

These data clearly indicate that a subtly balanced synthesisof the motB and motC gene products is absolutely required forcoordinated flagellar rotation. Consistent with this notion isthe motile behavior of a merodiploid (RU11/514) that carriesa chromosomal motB-motC deletion and the native motB-motC configuration on the complementing plasmid, pRU1902.This strain with an increased but balanced motB-motC geneexpression exhibited near-normal swarming (Fig. 5A) and free-swimming behavior (Table 2).

(ii) MotC. The derived MotC polypeptide (Fig. 3) has noknown precedents in other bacteria. The protein possesses a

FIG. 2. Physical and genetic map of a composite 7.7-kbp EcoRI fragment containing three genes, motB, motC, and motD, of the R. meliloti mot operon. Pointedopen boxes mark the polarity and extent of genes. Sites of Tn5 insertions inducing the jiggly (RU11/112) and paralyzed (RU11/113) phenotypes are indicated by verticalarrows. The extents of deletions introduced in motB (RU11/218), motC (RU11/211), motD (RU11/212), and motB-motC (pRU11/513), respectively, are indicated bytwo-pointed arrows. Hooked arrows define the 4.2-kbp sequenced portion shown in Fig. 3.

FIG. 3. Nucleotide sequence of a 4,196-bp R. meliloti genomic region (Fig. 2) containing the mot operon and deduced amino acid sequences. Numbers to the leftrun sequentially and refer to nucleotides (upper lines) and amino acid residues (lower lines). The 9-bp target sites of two Tn5 insertions are doubly underlined andmarked by mutant strain numbers. Presumptive promoter boxes (210, 235), transcription termination (4 z 3), ribosome-binding sites (SD), translation starts (geneassignments), and stops (p) are indicated. Dotted underlining marks the extent of deletions introduced to yield RU11/218 (DmotB), RU11/211 (DmotC), and RU11/212(DmotD); the large deletion RU11/513 (D[motB-motC]) is delimited by vertical lines. Six nucleotides (TGAGAA) deleted to fuse motB and motC (RU11/213) are boxed.Cloned fragments used for complementation analysis are delimited by hooked arrows with plasmid assignments (Table 1; Fig. 2). The vertical arrow marks the putativesignal peptidase cleavage site in MotC.



.org/D

ownloaded from

http://jb.asm.org/



.org/D

ownloaded from

http://jb.asm.org/

typical signal sequence for secretion into the periplasm (49)with a short, positively charged N-terminal sequence (MLKR)followed by 16 mostly hydrophobic residues (LCTLLAASALAAPLAL) and a signal peptidase I site (GLARA) with theconsensus GX(X)AXA. A hydropathy plot of MotC (28) re-vealed no additional membrane-spanning domains, suggestingthat the mature MotC protein is located in the periplasm.

This assertion was proven correct by Western blot analysis ofcrude cell extracts and cell fractions prepared from wild-type

and mutant R. meliloti (Fig. 6). A polyclonal antiserum raisedagainst recombinant MotC protein revealed a 53-kDa band,present in crude extracts and the periplasmic fraction (Fig. 6,lanes 1 and 4) but not in the cytoplasmic or membrane frac-tions (lanes 2 and 3) of wild-type RU11/001. The 53-kDa bandwas absent from a motC deletant (RU11/211 [lane 5]) butpresent in the periplasmic fraction if this strain carried a com-plementing motC gene (RU11/215 [lane 8]), thus proving thevalidity of this analysis. b-Galactosidase and alkaline phospha-tase served as marker enzymes of the cytoplasmic and periplas-mic fractions, respectively. The presence or absence of theseenzymes in the cell fractions tested clearly demonstrated thatthese were properly assigned to the respective bacterial cellularcompartments (data not shown).

Two Tn5 insertions differentially affect motC. One insertion,separating the motC codons 45 and 46 (RU11/113 [Fig. 3]),resulted in paralysis, whereas the other (RU11/112 [Fig. 3]), 5

FIG. 4. (A) Aligned polypeptide sequences of motB-encoded proteins from R. meliloti ([rm]), E. coli ([ec]) (64), and B. subtilis ([bs]) (43). An a-helical domain (p)putatively spanning the membrane and containing an aspartate residue (2) believed to participate in proton transfer is marked; conserved residues are highlighted byshading. A deduced peptidoglycan-binding domain is indicated by alignment with the consensus (X 5 any residue) (9, 26). Alignments were optimized by using PILEUPof the GCG program package (10).

FIG. 5. Effects of deletions and trans complementation of motB-motC to-gether and of motB, motC, and motD alone on swarming. See Table 1 fordescriptions of the strains tested. (A) 1, RU11/001 (wild type); 2, RU11/513(D[motB motC]); 3, RU11/514 (D[motB motC]/motB-motC). (B) 1, RU11/001; 2,RU11/218 (DmotB); 3, RU11/219 (DmotB/motB). (C) 1, RU11/001; 2, RU11/211(DmotC); 3, RU11/215 (DmotC/motC); 4, RU11/213 (motB-motC fusion). (D) 1,RU11/001; 2, RU11/212 (DmotD); 3, RU11/216 (DmotD/motD). Experimentalconditions were as for Fig. 1.

TABLE 2. Absolute and relative free-swimming speeds of wild-typeand mutant R. melilotia

Strain Vinst (mm s21)b Vsl (mm s21)c Vsl/Vinstd

RU11/001 (wild type) 32.8 6 0.4 25.8 6 0.5 0.79RU11/112

(Tn5::motC)20.5 6 0.6 10.9 6 1.1 0.53

RU11/219 (DmotB/motB)e

25.4 6 1.2 15.3 6 0.6 0.60

RU11/215 (DmotC/motC)e

26.8 6 1.4 12.3 6 2.3 0.45

RU11/514 (D[motBmotC]/motB motC)

31.6 6 0.7 23.0 6 1.0 0.73

RU11/216 (DmotD/motD)

35.9 6 0.9 26.0 6 1.1 0.73

a Mean values of 100 to 300 individual cell tracks were determined from eachsample by computerized motion analysis and were averaged from at least fiveindependent cell populations.

b Vinst, instantaneous velocity; defines the velocity measured in every trackpoint and averaged for the track (absolute speed).

c Vsl, straight-line velocity; defines the linear distance traveled per secondbetween endpoints of a given track (relative speed).

d Smooth-swimming rate; represents the proportion of smooth swimming un-der a given condition.

e About 5% normal-swimming cells.



.org/D

ownloaded from

http://jb.asm.org/

bp upstream of the motC translation start, resulted in jigglyswimming (Fig. 1). This insertion affects neither the TAG stopof motB nor the ATG start of motC, owing to a 9-bp sequenceduplication generated by Tn5, but interrupts continuous tran-scription and the translation coupling of motB and motC. Con-ceivably, the Tn5 insertion generates a weak promoter up-stream of motC that leads to low-level transcription andreduced synthesis of MotC relative to MotB.

Furthermore, a motB-motC gene fusion generated by dele-tion of 6 bp—the motB stop codon plus 3 bp between motB andmotC (Fig. 3)—resulted in a nonswarming mutant phenotype(RU11/213 [Fig. 5C]). When cells were grown in Bromfieldmedium, about 5% uncoordinated, jiggly-swimming cells wereobserved, whereas the majority of cells were paralyzed. Sup-pressor mutants that regained between 50 and 100% of wild-type swarming proficiency were isolated from swarm platesafter 10 to 12 days of incubation at 30°C. PCR-based sequenceanalyses revealed intragenic suppression in motB, either byC-T transition at position 2178 or by deletion of a G at position2176 or 2182 (Table 3). As a result, each suppressor mutanthad a new motB stop codon and produced MotB variants withtwo amino acids deleted or added at the C terminus. Theformer regained 100% and the latter regained 50 to 60% oftheir swarming proficiency. These results suggest that the in-tegrity of motB and motC genes is a prerequisite for normalmotor function.

(iii) MotD. The third open reading frame, motD, encodes a475-residue polypeptide. Database searches did not reveal anysimilarity with known proteins. The derived polypeptide con-tains no obvious membrane-spanning domains, nor has a signalpeptide been detected. Therefore, MotD is likely to be a cy-

toplasmic protein, with an essential function in motility. Adeletion of the 59 half of motD (RU11/212 [Fig. 2]) renderedthe cells paralyzed and nonswarming (Fig. 5D), though elec-tron microscopy revealed normal flagellation. In contrast toresults for motB and motC, complementation with an overex-pressed native motD gene on plasmid pRU2015 (strain RU11/213 [Table 1]) even overcompensated for the swarming defi-ciency of the deletion mutant RU11/212 (Fig. 5D), and free-swimming speed was increased by 10% (Table 2). The resultssuggest that an enhanced dose of MotD positively affectsflagellar rotary speed.

DISCUSSION

We have previously shown that R. meliloti responds to tacticstimuli by rotary speed control of its flagella (15, 60). Thefinding of two new mot genes, motC and motD, offers interest-ing possibilities for this new mode of flagellar rotation. Byanalysis of the encoded Mot proteins and various mutant phe-notypes, it may be possible to assign functions to these motilityproteins in a working model whose predictions can be tested.

Like in E. coli, the conserved MotA (62) and MotB (Fig. 4)proteins of R. meliloti are considered essential for energizingthe flagellar motor. Four predicted membrane-spanning a he-lices of MotA and a single a helix of MotB are thought to forma proton-conducting channel (3–5). Based on topological dataand sequence comparisons, MotB has been proposed as linkerconnecting the MotA-MotB stator complex to the cell wall (7,9, 26). An analysis of unusual speed fluctuations in slow-swim-ming mutants of E. coli (mapped to the large periplasmicdomain of MotB) indicated that in these mutants, the numberof force-generating units was reduced (5). This observation isinteresting in view of a possible mechanism for flagellar rotaryspeed variation that inherently directs the swimming pattern ofR. meliloti cells (60).

The motB and motC genes of R. meliloti are closely linked ina way that implies coordinate expression and stoichiometricrelationships between the gene products, MotB and MotC.This view is supported by the properties of mutants that alterthe balance of transcription (strain RU11/112) or translation(strain RU11/213) and that either under- or overexpress MotCprotein relative to MotB (strains RU11/215 and RU11/219[Table 1]). Their jiggly swimming patterns reflect uncoordi-nated rotation of different flagella on a given cell, as evidencedby extreme fluctuations of rotary speed observed in tetheredmutant cells (data not shown). Therefore, the correct stoichi-ometry and positioning of MotC relative to MotB, which isfaulty in the abnormal swimmers, may be the key to controllingthe rotary speed of a flagellum. The need for balanced biosyn-thesis of MotB and MotC may be explained by the sequence ofevents that requires the correct placing of MotB first to scav-enge and position MotC correctly at the periplasmic surface ofthe flagellar motor (see below).

FIG. 6. Western blot analysis of crude extracts and of cell fractions fromwild-type and mutant R. meliloti, using polyclonal anti-MotC antiserum raisedagainst reMotC protein. Cell fractionation and visualization of bands by chemi-luminescence are described in Materials and Methods. Arrow marks the MotCband with an apparent molecular weight of 53,000. Lanes 1 to 4, fractions fromRU11/001 (wild type) (cell extract, membranes, cytoplasm, and periplasm, re-spectively); lane 5, RU11/211 (DmotC), cell extract; lanes 6 to 8, fractions fromRU11/215 (DmotC/motC) (membranes, cytoplasm, and periplasm, respectively).Positions of molecular mass standards are indicated to the left in kilodaltons.

TABLE 3. Nucleotide changes, corresponding amino acid changes, and the effect on swarming in suppressor mutants ofRU11/213 (motB-motC fusion)

motB allele(s) Nucleotide change Amino acid change Swarming proficiencya

(% of wild type)

213 DT2184–A2189 Fusion MotB-MotC 5500, 505, 506 C21783T Gln3933stop 100504 DG2176 Gly392 Gln393 Gly3943Ala

Arg Ala Cys stop50

508 DG2182 Gly3943Ala Cys stop 40

a Defined as the ratio of diameter of mutant swarm ring to diameter of wild-type swarm ring 3 100 on a Bromfield swarm plate.



.org/D

ownloaded from

http://jb.asm.org/

The N-terminal amino acid sequence of the derived MotCpolypeptide has features of a signal sequence (49) suggestingthat MotC is secreted to the periplasm. This notion was provencorrect by Western analysis of cell fractions that located MotCin the periplasm of the R. meliloti cell (Fig. 6). The antiserum-complexed MotC band was detectable in neither the cytoplas-mic nor membrane fraction of wild-type R. meliloti, and it wascompletely lacking in the motC deletion strain RU11/211. Thelikely partner of a periplasmic MotC protein that exerts controlover motor speed would be MotB protein with its largeperiplasmic domain (7). In fact, preliminary data obtained withcloned portions of motB and motC in the yeast two-hybridsystem (12) indicate protein-protein interactions betweenMotC and the periplasmic portion of MotB (17).

These findings together with the properties of certain slow-swimming E. coli motB mutants (5, 65) are consistent with aworking model that accounts for flagellar rotary speed controlexerted by MotB-MotC interactions (Fig. 7). The model isbased on the assumptions (i) that in R. meliloti, fluctuations ofrotary speed are a function of the fraction of working MotA-MotB complexes (force-generating units or torque generators)acting on a given motor (24) and (ii) that interactions betweenMotB and MotC control the activity of individual force-gener-ating units. MotB alone leaves a torque generator nonfunc-tional, but it becomes functional when MotB and MotC arecomplexed. A missing or defective MotC molecule thus arrestsproton conduction or energy coupling of a force-generatingMotA-MotB complex, a deficit that reduces the rotary speed ofthat motor. This model is consistent with the observed abnor-mal (jiggly) swimming pattern of mutants with an upset bal-ance of MotB and MotC proteins (strains RU11/112, RU11/213, and RU11/215 [Fig. 5 and Table 2]), whose swimmingbehavior results from rotary speed fluctuations that preventflagellar synchrony in a single cell.

MotD, the other new motility protein, has features of acytoplasmic protein thought to be associated with the innersurface of the motor (Fig. 7). Complementation studies suggestthat overexpression of motD even improves normal swarmingor free-swimming motility (Fig. 5D and Table 2). MotD may bea cytoplasmic relay involved in transmitting signals from theresponse regulators, CheY1 and CheY2 (57), to the speed-controlling elements. If our current experiments (17) confirmthe alleged role of MotB-MotC interactions in motor speedcontrol, an elucidation of the signal pathway from cytoplasm toperiplasm will be the topic of future studies.

ACKNOWLEDGMENTS

This work was supported by the Deutsche Forschungsgemeinschaft(Schm68/24-2) and by a DAAD travel grant (ARC).

We thank Judy Armitage for helpful discussion, Simon Silver forcritical review of the manuscript, and Klaus Stark for artwork.

REFERENCES

1. Bachmann, B. J. 1990. Linkage map of Escherichia coli K-12, edition 8.Microbiol. Rev. 54:130–197.

2. Blair, D. F., and H. C. Berg. 1988. Restoration of torque in defective flagellarmotors. Science 242:1678–1681.

3. Blair, D. F., and H. C. Berg. 1990. The MotA protein of E. coli is a proton-conducting component of the flagellar motor. Cell 60:439–449.

4. Blair, D. F., and H. C. Berg. 1991. Mutations in the MotA protein ofEscherichia coli reveal domains critical for proton conduction. J. Mol. Biol.221:1433–1442.

5. Blair, D. F., D. Y. Kim, and H. C. Berg. 1991. Mutant MotB proteins inEscherichia coli. J. Bacteriol. 173:4049–4055.

6. Brickman, E., and J. Beckwith. 1975. Analysis of the regulation of E. colialkaline phosphatase synthesis using deletions and 80 transducing phages. J.Mol. Biol. 96:307–316.

7. Chun, S. Y., and J. S. Parkinson. 1988. Bacterial motility: membrane topol-ogy of the Escherichia coli MotB protein. Science 239:276–278.

8. Dean, G. E., R. M. Macnab, J. Stader, P. Matsumura, and C. Burks. 1984.Gene sequence and predicted amino acid sequence of the motA protein, amembrane-associated protein required for flagellar rotation in Escherichiacoli. J. Bacteriol. 159:991–999.

9. De Mot, R., and J. Vanderleyden. 1994. The C-terminal sequence conserva-tion between OmpA-related outer membrane proteins and MotB suggests acommon function in both Gram-positive and Gram-negative bacteria, pos-sibly in the interaction of these domains with peptidoglycan. Mol. Microbiol.12:333–334.

10. Devereux, J., P. Haeberli, and O. Smithies. 1984. A comprehensive set ofsequence analysis programs for the VAX. Nucleic Acids Res. 12:387–395.

11. Dietmaier, W., S. Fabry, and R. Schmitt. 1993. DISEC-TRISEC: di- andtrinucleotide-sticky-end cloning of PCR-amplified DNA. Nucleic Acids Res.21:3603–3604.

12. Fields, S., and O. Song. 1989. A novel genetic system to detect protein-protein interactions. Nature 340:240–241.

13. Francis, N. R., G. E. Sosinsky, D. Thomas, and D. J. DeRosier. 1994.Isolation, characterisation and structure of bacterial flagellar motors con-taining the switch complex. J. Mol. Biol. 235:1261–1270.

14. Gotz, R., N. Limmer, K. Ober, and R. Schmitt. 1982. Motility and chemotaxisin two strains of Rhizobium with complex flagella. J. Gen. Microbiol. 128:789–798.

15. Gotz, R., and R. Schmitt. 1987. Rhizobium meliloti swims by unidirectional,intermittent rotation of right-handed flagellar helices. J. Bacteriol. 169:3146–3150.

16. Greck, M., J. Platzer, V. Sourjik, and R. Schmitt. 1995. Analysis of a che-motaxis operon in Rhizobium meliloti. Mol. Microbiol. 15:989–1000.

17. Haslbeck, M., W. Sterr, and R. Schmitt. Unpublished data.18. Helmann, J. D. 1991. Alternative sigma factors and the regulation of flagellar

gene expression. Mol. Microbiol. 5:2875–2882.19. Higuchi, R. 1989. Using PCR to engineer DNA, p. 61–70. In H. A. Erlich

FIG. 7. Working model of the R. meliloti flagellar motor. The cartoon illustrates what is known about the basal body structure and its relation to layers of the cellenvelope of gram-negative bacteria (13, 35). (A) The suggested interaction of MotC with the periplasmatic domain of MotB implicated in proton flux and thepresumptive location of MotD at the cytoplasmic surface of the motor are depicted. We assume this to be the active configuration with a functional force-generatingMotA-MotB complex. (B) Possible effect of a mutation or deletion in motC leading to a defective MotCp protein. The failure to bind to MotB is thought to result inloss of energy flux, as schematically indicated by distortion of the MotA-MotB complex.



.org/D

ownloaded from

http://jb.asm.org/

(ed.), PCR technology. Principles and applications for DNA amplification.Stockton Press, New York, N.Y.

20. Hirota, N., and Y. Imae. 1983. Na1-driven flagellar motors of an alkalophilicBacillus strain YN-1. J. Mol. Biol. Chem. 258:10557–10581.

21. Irikura, V. M., M. Kihara, S. Yamaguchi, H. Sockett, and R. M. Macnab.1993. Salmonella typhimurium fliG and fliN mutations causing defects inassembly, rotation, and switching of the flagellar motor. J. Bacteriol. 175:802–810.

22. Jones, C. J., and S. Aizawa. 1991. The bacterial flagellum and flagellarmotor: structure, assembly and function. Adv. Microb. Physiol. 32:109–172.

23. Kamberger, W. 1979. An Ouchterlony double diffusion study on the inter-action between legume lectins and rhizobial cell surface antigens. Arch.Microbiol. 121:83–90.

24. Khan, S., M. Dapice, and T. H. Reese. 1988. Effects of mot gene expressionon the structure of the flagellar motor. J. Mol. Biol. 202:575–584.

25. Kihara, M., M. Homma, K. Kutsukake, and R. M. Macnab. 1989. Flagellarswitch of Salmonella typhimurium: gene sequences and deduced proteinsequences. J. Bacteriol. 171:3247–3257.

26. Koebnik, R. 1995. Proposal for a peptidoglycan-associating alpha-helicalmotif in the C-terminal regions of some bacterial cell-surface proteins. Mol.Microbiol. 16:1269–1270.

27. Krupski, G., R. Gotz, K. Ober, E. Pleier, and R. Schmitt. 1985. Structure ofcomplex flagellar filaments in Rhizobium meliloti. J. Bacteriol. 162:361–366.

28. Kyte, J., and R. F. Doolittle. 1982. A simple method for displaying thehydropathic character of a protein. J. Mol. Biol. 157:105–132.

29. Labes, M., A. Puhler, and R. Simon. 1990. A new family of RSF1010-derivedexpression and lac-fusion broad-host-range vectors for gram-negative bacte-ria. Gene 89:37–46.

30. Larsen, S. H., R. W. Reader, E. N. Kort, W.-W. Tso, and J. Adler. 1974.Change in direction of flagellar rotation is the basis of the chemotacticresponse in Escherichia coli. Nature 249:74–77.

31. Lehrer, R. J., A. Barton, and T. Ganz. 1988. Concurrent assessment of innerand outer membrane permeabilization and bacteriolysis in E. coli by multi-ple-wave length spectrophotometry. J. Immunol. Methods 108:153–158.

32. Li, M., and H. P. Schweizer. 1993. Resolution of common DNA sequencingambiguities of GC-rich DNA templates by terminal deoxynucleotidyl trans-ferase without dGTP analogues. Focus 14/15:19–20.

33. Long, S., S. McCune, and G. C. Walker. 1988. Symbiotic loci of Rhizobiummeliloti identified by random TnphoA mutagenesis. J. Bacteriol. 170:4257–4265.

34. Luria, S. E., F. N. Adams, and R. C. Ting. 1960. Transduction of lactoseutilizing ability among strains of E. coli and S. dysenteriae and the propertiesof the transducing phage particles. Virology 12:348–390.

35. Macnab, R. M. 1996. Flagella and motility, p. 123–145. In F. C. Neidhardt,R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S.Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichiacoli and Salmonella: cellular and molecular biology, 2nd ed. American So-ciety for Microbiology, Washington, D.C.

36. Malakooti, J., Y. Komeda, and P. Matsumara. 1989. DNA sequence analysis,gene product identification, and localization of flagellar motor componentsof Escherichia coli. J. Bacteriol. 171:2728–2734.

37. Manson, M. D., P. Tedesco, H. C. Berg, F. M. Harold, and C. van der Driftc.1977. A proton motive force drives bacterial flagella. Proc. Natl. Acad. Sci.USA 74:3060–3064.

38. Marsh, J. L., M. Erfle, and E. J. Wykes. 1984. The plC plasmid and phagevectors with versatile cloning sites for recombinant selection by insertionalinactivation. Gene 32:482–485.

39. Martin, J. L., G. Waksman, J. C. A. Bardwell, J. Beckwith, and J. Kuriyan.1993. Crystallization of DsbA, an E. coli protein required for disulphide bandformation in vivo. J. Mol. Biol. 230:1097–1100.

40. Maruyama, M., G. Lodderstaedt, and R. Schmitt. 1978. Purification andbiochemical properties of complex flagella isolated from Rhizobium lupiniH13-3. Biochim. Biophys. Acta 535:110–124.

41. Matsuura, S., J. Shioi, and Y. Imae. 1977. Motility in Bacillus subtilis drivenby an artificial proton motive force. FEBS Lett. 82:187–190.

42. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring HarborLaboratory Press, Cold Spring Harbor, N.Y.

43. Mirel, D. B., V. M. Lustre, and M. J. Chamberlin. 1992. An operon ofBacillus subtilis motility genes transcribed by the sigma-D form of RNApolymerase. J. Bacteriol. 174:4197–4204.

44. Novick, R. P., R. C. Clowes, S. N. Cohen, R. Curtis, N. Datta, and S. Falkow.1976. Uniform nomenclature for bacterial plasmids: a proposal. Bacteriol.Rev. 40:168–189.

45. Olsen, G. J., C. R. Woese, and R. Overbeek. 1994. The winds of (evolution-ary) change: breathing new life into microbiology. J. Bacteriol. 176:1–6.

46. Oppenheim, D. S., and C. Yanofsky. 1980. Translational coupling duringexpression of the tryptophan operon of Escherichia coli. Genetics 95:785–795.

47. Pleier, E., and R. Schmitt. 1989. Identification and sequence analysis of tworelated flagellin genes in Rhizobium meliloti. J. Bacteriol. 173:1467–1476.

48. Pleier, E., and R. Schmitt. 1991. Expression of two Rhizobium meliloti genesand their contribution to the complex filament structure. J. Bacteriol. 173:2077–2085.

49. Pugsley, A. P. 1993. The complete general secretory pathway in gram-neg-ative bacteria. Microbiol. Rev. 57:50–108.

50. Rosenberg, A. H., B. N. Lade, D. Chui, S. Lin, J. J. Dunn, and F. W. Studier.1987. Vectors for selective expression of cloned DNAs by T7 RNA polymer-ase. Gene 56:125–135.

51. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, N.Y.

52. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing withchain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463–5467.

53. Schafer, A., A. Tauch, W. Jager, J. Kalinowski, G. Thierbach, and A. Puhler.1994. Small mobilizable multi-purpose cloning vectors derived from theEscherichia coli plasmids pK18 and pK19: selection of defined deletions inthe chromosome of Corynebacterium glutamicum. Gene 145:69–73.

54. Schuster, S. C., and S. Khan. 1994. The bacterial flagellar motor. Annu. Rev.Biophys. Biomol. Struct. 23:509–539.

55. Selbitschka, W., S. Niemann, and A. Puhler. 1993. Construction of genereplacement vectors for Gram2 bacteria using a genetically modified sacRBgene as positive selection marker. Appl. Microbiol. Biotechnol. 38:615–618.

56. Shao, Z., J. Platzer, and R. Schmitt. Unpublished data.57. Shine, J., and L. Dalgarno. 1974. The 39 terminal sequence of Escherichia

coli 16S ribosomal RNA: complementary to nonsense triplets and ribosomebinding sites. Proc. Natl. Acad. Sci. USA 71:1342–1346.

58. Simon, R., M. O’Connell, M. Labes, and A. Puhler. 1986. Plasmid vectors forthe genetic analysis and manipulation of rhizobia and other Gram-negativebacteria. Methods Enzymol. 118:640–659.

59. Sockett, H., S. Yamaguchi, M. Kihara, V. M. Irikura, and R. M. Macnab.1992. Molecular analysis of the flagellar switch protein FliM of Salmonellatyphimurium. J. Bacteriol. 174:793–806.

60. Sourjik, V., and R. Schmitt. 1996. Different roles of CheY1 and CheY2 inthe chemotaxis of Rhizobium meliloti. Mol. Microbiol. 22:427–436.

61. Stader, J., P. Matsumura, D. Vacante, G. E. Dean, and R. M. Macnab. 1986.Nucleotide sequence of the Escherichia coli motB gene and site-limitedincorporation of its product into the cytoplasmic membrane. J. Bacteriol.166:244–252.

62. Sterr, W. 1995. Diploma thesis. University of Regensburg, Regensburg,Germany.

63. Stolz, B., and H. C. Berg. 1991. Evidence for interactions between MotA andMotB, torque-generating elements of the flagellar motor of Escherichia coli.J. Bacteriol. 173:7033–7037.

64. Studier, F. W., and B. A. Moffat. 1986. Use of bacterial T7 RNA polymeraseto direct selective high-level expression of cloned genes. J. Mol. Biol. 189:113–130.

65. Togashi, F., S. Yamaguchi, M. Kihara, S.-I. Aizawa, and R. M. Macnab.1997. An extreme clockwise switch bias mutation in fliG of Salmonellatyphimurium and its suppression by slow-motile mutations in motA and motB.J. Bacteriol. 179:2994–3003.

66. Trachtenberg, S., D. J. De Rosier, S.-I. Aizawa, and R. M. Macnab. 1987.Three-dimensional structure of the complex flagellar filament of Rhizobiumlupini and its relation to the structure of the plain filament. J. Mol. Biol.195:603–620.

67. Wilson, M. L., and R. M. Macnab. 1990. Co-overproduction and localizationof the Escherichia coli motility proteins MotA and MotB. J. Bacteriol. 172:3932–3939.

68. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phagecloning vectors and host strains: nucleotide sequences of the M13mp18 andpUC19 vectors. Gene 33:103–119.



.org/D

ownloaded from

http://jb.asm.org/

three genes of a motility operon and their role in ... · a motility (mot) operon containing three...

Documents